Unstructured regridder : usage tidying #2273
Conversation
pp-mo
changed the title
pp-mo
force-pushed
the
unstructured_regridder_usagetidy
branch
from
63d639f to
bb712e4
Compare
marqh
left a comment
marqh
left a comment
There was a problem hiding this comment.
broadly supportive, a few comments and queries to look at please
| same units in the source and grid cubes. | ||
| .. Note:: | ||
| Currently only supports regridding, not interpolation. |
There was a problem hiding this comment.
This distinction isn't immediately obvious to me: interpolation is a numerical method which can support a regrid operation.
As this is the docstring for a class called UnstructuredNearest I suggest that this note is simply not required
There was a problem hiding this comment.
This distinction isn't immediately obvious to me: interpolation is a numerical method which can support a regrid operation.
What the words 'regrid' and 'interpolate' mean in Iris terms is that cubes support both a 'regrid' and an 'interpolate' method, which can both specify a "scheme".
the docstring for a class called UnstructuredNearest
UnstructuredNearest is a scheme, and these can generally be used with either interpolate or regrid
-- except that this one can't. ( Similarly, neither can PointInCell ).
So, I'd like to keep this statement.
This arrangement is now well established in Iris.
The practical difference is that a "regrid" uses a grid cube as its reference, while "interpolate" takes coordinate values.
But I do agree that the need for both is questionable, and the choice between is nowhere very clearly explained !!
| .. Note:: | ||
| If the points are longitudes/latitudes, these are handled correctly as | ||
| points on the sphere, but the values must be in 'degrees'. |
There was a problem hiding this comment.
is this protected somewhere? if i supply radians, will a meaningful exception be raised?
There was a problem hiding this comment.
is this protected somewhere? if i supply radians, will a meaningful exception be raised?
No, it is not checked and will just give wrong answers !
I decided to document this problem rather than fix it, as it is an existing low-level routine.
As this is a private routine I thought that was ok.
Arguably, we should maybe fix the existing analysis.trajectory.interpolate, as it calls this and so exposes the same problem.
? what do you think @marqh ?
I have fixed it in the UnstructuredNearest classes, which now convert everything to degrees (and complain if it fails).
| * src_cube: | ||
| The :class:`~iris.cube.Cube` defining the source grid. | ||
| The X and Y coordinates must be mapped over the same dimensions. | ||
| The X and Y coordinates can have any shape, but must be mapped over |
| def test_nearest(self): | ||
| res = self.src.regrid(self.grid, UnstructuredNearest()) | ||
| self.assertArrayShapeStats(res, (1, 6, 3, 4), 315.888737, 11.000729) | ||
| self.assertArrayShapeStats(res, (1, 6, 3, 4), 315.890808, 11.000724) |
There was a problem hiding this comment.
are these changes expected?
the value differences are small, but non-zero
There was a problem hiding this comment.
The operation result was previously actually wrong, because the test data has coords with units of 'radians' (!)
The change is small, I think because all the source points (temperatures) actually have quite similar values.
It just goes to show you have to be careful with statistical tests.
indeed |
|
it seems there's no outstanding issues here, I am go for launch cheers @pp-mo |
2 participants
A few functional changes / fixes.
A lot of documentation and testing enhancements.